This is a renewal of a project that has enrolled 707 subjects, including 527 depressed elderly patients and 180 non-depressed elderly controls over the past ten years. Over the next five years we will follow 254 active subjects, including 182 patients and 72 controls, with a focus on genetic and post-mortem neuroanatomical correlates of mood and cognition. These subjects are currently being followed in the current study, and no new subjects will be added. Analyses will be performed on all 757 subjects who have been enrolled. Detailed psychosocial, functional, clinical, psychiatric, medical, neurological, and cognitive assessments will continue to be obtained at defined points during follow-up. At least two brain MRI studies will have been performed on the majority of subjects at a two year interval; the existing MRI data will be used in longitudinal analyses as well. The principal outcome measures are trajectory of mood and cognition and on post-mortem neuroanatomical and neuropathological brain changes. The analysis plan focuses on examination of risk for recurrence and chronicity of depression and risk for cognitive decline and later dementia using the following independent variables: baseline cognitive profile, baseline fractional anisotropy (FA) and change in FA, functional MRI measure of resting state connectivity, and depression course. The project will preserve past methodological advances by combining psychiatric assessments with psychosocial and psychobiological perspectives. In a study design that employs carefully defined treatment algorithms, we will test specific hypothesis regarding mood outcomes and cognitive decline and dementia. A novel feature of the study is examination of post-mortem brain changes, specifically alterations in basolateral amygdala. It is expected that the results from this study will clarify the relationship between depression and dementia in the elderly. It will also add to the literature on neuroimaging factors and the long-term outcome of depression in a clinical setting. It is the first prospective longitudinal study in depression to correlate clinical in vivo data on well characterized patients with subsequent post-mortem morphometric analyses.